This study will use the genome network as a framework for predicting evolution in a virus. The viral genome will be modified in simple ways by deleting or replacing a single gene, and the modified virus will be adapted to observe how it evolves in response to the modification. The general prediction is that changes which evolve in the modified virus will occur in the parts of the genome known to interact with the modified gene. The study will provide a direct test of this framework with experimental data. The virus used is the harmless bacteriophage T7, which offers a good model system for other viruses because much is already known about T7 molecular biology, a genome network can be assembled, and T7 grows only on a few strains of bacteria. The work will give useful insight to several questions and goals in public health and molecular epidemiology, as well as in basic biology: (i) the feasibility of engineering new genomes and creating recombinant vaccines safely; (ii) how well we can predict molecular evolution; (iii) whether genome interactions are conserved across related species; (iv) how to detect genetic signatures of evolution in engineered versus natural genomes. [unreadable] [unreadable] [unreadable]